On Detection Strategies for Linearly Precoded MIMO-OFDM Systems with CSIT

This paper explores in detail the application of optimum detection based on maximum likelihood principles in linearly precoded subcarrier cooperative MIMO-OFDM systems. Two different strategies, full spreading and efficient spreading, are presented to disperse the information in the frequency and space domains with the objective of increasing the diversity gain of the system under the assumption of optimum detection at the receiver. The application of efficient spreading at the transmitter allows the implementation of a low-complexity optimal detector that combines linear and non-linear processing. These two techniques, full and efficient spreading, represent different operating points in the performance-complexity plane as shown by the analytical derivation of their diversity order. Remarkably, this analysis serves to reveal and quantify the benefits of subcarrier cooperation with optimum reception. Finally, and in order to highlight the benefits of the proposed approach in practical setups such as those found in contemporary wireless standards, the receiver architectures are extended to operate with soft information, thus allowing the use of iterative soft based receivers. Simulation results demonstrate that efficient spreading in combination with iterative decoding is an attractive combination enabling the implementation of high throughput wireless systems with a large degree of robustness.